The purpose of this project is to elucidate the principles of treatment of acute and chronic pain syndromes, with particular attention to the drug treatment of pain caused by nerve injury and surgery. In randomized, double-blind, crossover studies in 84 patients with painful diabetic neuropathy, 74% of patients had moderate or better relief with amitriptyline (an antidepressant blocking both norepinephrine and serotonin reuptake), 61% with the selective norepinephrine reuptake blocker desipramine, 48% with the selective serotonin reuptake blocker fluoxetine, and 41% with placebo. These results suggest that desipramine is a useful alternative in patients unable to tolerate amitriptyline side effects, that fluoxetine has little effect at the dose tested, and that norepinephrine may be a mediator of the analgesia produced by tricyclic antidepressants. In a study of the noradrenergic agonist clonidine, administered by transdermal patch, 54% of 24 patients obtained moderate relief or better, but pain scores were not significantly lower than recorded during placebo treatment. A subset of 8 patients had a marked analgesic response to clonidine and not placebo, and in many of these that differential response has been replicated. Clonidine may be an effective treatment for a subset of patients with neuropathic pain, but studies need to be designed to maximize the power to detect response in a subset. In patients with pain following orthopedic or gynecological surgery, a single does of desipramine, 50 mg, did not potentiate the effects of intravenous morphine unlike results in animals and in a small trial at another center that used chronic desipramine treatment. Future studies of this possible interaction should use either single doses higher than 50 mg or chronic treatment.